Measuring Co-Skewness and Co-kurtosis using Noisy High Frequency Financial Data
Zhi Liu
Department of Mathematics, University of Macau, Macau, China
Junwei Liu
The Wangyanan Institute for Studies in Economics, Xiamen University, China
Kent Wang
8/19/2013 11:55:57 PM
Few of works in literature have paid attention to the higher co-moments of asset return under high frequency situation. In this paper, we present a new measure of higher co-moments, i.e., co-skewness and co-kurtosis using ultra high frequency nancial data under the general jump di usion model. Moreover, the estimator is robust to the presences market microstructure noise, which usually appears in ultra high frequency data. The consistency is proved. The simulation studies con rm the performance of the estimator. We also implement the new measure to the real high frequency nancial data, documenting a strong time series pricing ability. The new measure has potential application in portfolio selection.
It^o semi-martingale; High frequency data; Market microstructure; Microstructure noise; Co-skewness; Co-Kurtosis; Pre-averaging.